发现报告（www.fxbaogao.com）汇聚了全网最全的研报资源，数量之多令人惊叹。作为金融圈用户量极大的平台，我们覆盖了宏观、行业、公司及财报等全领域。我们追求极致的效率，设计极简，技术领先，帮您从海量数据中迅速抓取重点。无论是行业研究还是投资分析，这里都是您高效获取信息、实现精准决策的得力工具。
This paper presents an experimental study on the use of ChatGPT for robotics applications. The authors outline a strategy that combines design principles for prompt engineering and the creation of a high-level function library to allow ChatGPT to adapt to different robotics tasks, simulators, and form factors. The study focuses on the effectiveness of different prompt engineering techniques and dialog strategies towards the execution of various types of robotics tasks. The authors explore ChatGPT's ability to use free-form dialog, parse XML tags, and synthesize code, in addition to the use of task-specific prompting functions and closed-loop reasoning through dialogues. The study encompasses a range of tasks within the robotics domain, from basic logical, geometrical, and mathematical reasoning to complex domains such as aerial navigation, manipulation, and embodied agents. The authors show that ChatGPT can be effective at solving several of such tasks, while allowing users to interact with it primarily via natural language instructions. In addition to these studies, the authors introduce an open-sourced research tool called PromptCraft, which contains a platform where researchers can collaboratively upload and vote on examples of good prompting schemes for robotics applications, as well as a sample robotics simulator with ChatGPT integration, making it easier for users to get started with using ChatGPT for robotics.
